acceleration of reduction of natality in Europe, proceeding from Sweden (western Europe) to Hungary (central Europe) and to Bulgaria (eastern Europe); (b) the somewhat slower declines in countries where the downward trend began at lower levels of natality (Japan and Puerto Rico), (c) the precipitous decline characteristic of some east Asian countries that have entered the transition since World War II (Taiwan).
If the experience so far is indicative of what may occur in other less developed countries the situation is more hopeful than often described. United Nations projections assume a 5 percent reduction during the first 5 years of sustained fertility decline, followed by about 10 percent for each subsequent 5-year period (20). These projections are conservative as compared with recent experience of countries undergoing transition.
As in the case of mortality, it would appear that a new pattern of fertility change is emerging in quite a few countries of the less developed world. The comparatively slow rate of natality reduction experienced in western Europe is being replaced by a new pattern of precipitous declines.
Recent data suggest that, once clearly begun, sustained fertility declines will proceed more rapidly in the less developed regions than they did historically in the West. But when such fertility declines will occur, and at -what level of socioeconomic development are, of course, critical factors. Is there a consistent threshold of socioeconomic development at which sustained reduction in fertility may be expected to occur? Does it differ from one major cultural region to another?
RELATIONSHIP BETWEEN SOCIOECONOMIC DEVELOPMENT AND REDUCTION OF THE BIRTH RATE
The so-called "threshold" hypothesis was examined in some detail for the United Nations study of world natality as of about the year 1960 (6, pp. 148-151). This study was unsuccessful in finding specific threshold values for various socioeconomic indicators, taking all the countries of the less developed world as a single group. However, when the analysis is confined to each of the several major cultural regions, more consistent patterns emerge.
In Latin America there is substantial correlation between major indices of development and natality. This is illustrated by data for the Latin American region in Table 4.* The indices include measures of urbanization, of economic structure (percent of economically active males in agriculture), of ecluca-
*The materials on Latin America are drawn from an unpublished study by K. S. Srikantan and the author of this paper.ropean because of its quite different sociocconomic characteristics and demographic behavior.